ArabCulture-Dialogue - MCQ (Dialect): leaderboard
Metric: Accuracy (%) selecting the one culturally appropriate continuation of three for each dialogue of ArabCulture-Dialogue (13 Arab countries, parallel Modern Standard Arabic and country-dialect versions of 3,471 culturally grounded multi-turn dialogues built from ArabCulture), dialect dialogues, no location context, scored by likelihood of the options, single run; chance is 33.3; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 94.8 |
| 2 | Gemini 2.5 Pro | 93.9 |
| 3 | Hala-9B | 73.3 |
| 4 | SILMA-9B-Instruct-v1.0 | 71.6 |
| 5 | Jais-2-8B-Chat | 71 |
| 6 | Gemma 2 9B (IT) | 60.9 |
| 7 | jais-adapted-7B-chat | 49.8 |
| 8 | Llama 3.1 8B Instruct | 41.2 |
| 9 | ALLaM-7B-Instruct-preview | 39.8 |
| 10 | Qwen 3 8B | 35.4 |
| 11 | Fanar-1-9B | 35 |
| 12 | c4ai-command-r7B-arabic-02-2025 | 34.2 |
Interactive version: theaggregate.ai/benchmark?slug=arabculture-dialogue-mcq-dialect · How It Works · Data refreshed daily, snapshot 2026-10-07.